Authenticity of tasks in online ESL/EFL learning to foster transfer

Abstract Authentic learning tasks, resembling life-like situations, have been effectively used in this research with Task-based Language Teaching (TBLT) of English as a Second/Foreign Language (ESL/EFL). The present study also focuses on the implementation of the Content and Language Integrated Learning (CLIL) approach in ESL/EFL teaching comparing two online courses, a course containing more “traditional” tasks with one consisting of less authentic tasks. For more meaningful learning, a sufficient level of authenticity to enable holistic teaching is required. Towards this goal, the study put into effect two e-courses, with 50 participants, differing distinctly in levels of authenticity. The Method section describes the three instruments, used to collect the relevant data, which will provide answers to the equally posed three research questions. Specifically, the Integrative Learning Value (ILV) instrument measured knowledge transfer, the Approaches to Teaching Inventory (ATI) students’ learning performance in relation to more or less teacher/student-centered teaching strategies as well as the 5-Dimensional Framework for Authenticity (5DF) instrument for student perceived authenticity. Finally, results are presented and conclusions are discussed.


Introduction
Authentic learning in or out of class (e.g., online) has been defined by Gilmore (2007) as learning activities that resemble real-world events as closely as possible, and at the same time provides learners with usable and meaningful knowledge that can be applied in practical contexts. Herrington et al. (2003) provide a number of characteristics that define authentic learning activity. The concepts "authenticity" and "meaningfulness" are closely entwined in authentic learning. A number of researchers like Egbert (2020), Jacobs et al. (2022), Vellanki and Bandu (2021), and Purnawarman and Dajarati (2020) have argued that meaningful learning needs to facilitate the application of what is learned in an authentic context, in order for learners to become aware or even experience the practical purpose or relevance of what they had to learn in class. Authentic contexts thus provide learners with learning activities outside the classroom context, which will make them conscientious of what knowledge and skills are useful for in future work environments. Various studies (Buendgens-Kosten, 2014;García-Pinar, 2019;Gilmore, 2007;Jaccomard & Kuuse, 2016;Oura, 2001;Ozverir et al., 2017;Richards, 2001) revealed that the provision of more authentic learning tasks can have positive effects on learner motivation, can provide adequate cultural information about the L2, as well as assist students in effectively aligning L2 learning content with real contexts.
For research on online learning environments, the concept of authenticity can be expected to become even more essential in the near future. Much of "emergency remote teaching" after the outbreak of the Corona pandemic simply transformed subject matter to an online format and did not provide our students with more meaningful instruction through activating actual learning environments (Hodges et al., 2020). According to Ahmadi (2019), authenticity in language education can be considered as the interplay between language users, their learning contexts (learning tasks, learning environment), and the language discourse itself (text types). More authentic tasks turn learning contexts into more meaningful ones when learners are able to find purpose (or relevance) and value in them. To further illustrate this point, as mentioned in Reinhardt (2020), the example of teaching a text type as opinion/discussion essay could be supplemented with technology-enhanced learning that provides authentic learning opportunities. Such learning tasks could be performed on e-platforms while using Google Docs and web 2.0 tools, like H5P, Kahoot or Padlet. To then stimulate collaborative discussion in such a learning environment, we could ask learners to create a PowerPoint presentation together, by using e-platforms like Zoom, WebEx, MS Teams, Google Meet, Blackboard on their laptop, tablet, or smartphone. In such ways, students that engage in authentic learning are also provided with opportunities to interact with peers and teachers.
Learning effects of authentic tasks have been identified on increased student engagement, growth of existing knowledge, as well as on professional skills and critical thinking (Gilmore, 2007;Herrington & Kervin, 2007;Larsen et al., 2017). According to Radović, Firssova, et al. (2020) andPerkins and Salomon (1988Salomon ( , 1989Salomon ( , 1992Salomon ( , 2012, Wang, 2012 the instructional design for effective teaching should include a sufficient level of authenticity in tasks, so as for more experiential and meaningful learning to be fostered through the resemblance of learning to reallife events as well as the occurrence of knowledge transfer in different contexts. Such authentic learning tasks enable students to integrate their current with past experiences as well as understand novel ideas. Authentic learning in ESL/EFL teaching plays a crucial role in that it assists students to achieve knowledge transfer (Perkins & Salomon, 1988Salomon & Perkins, 1989) between contexts as well as make effective connections. Oliver et al. (2002) have identified nine elements of such authentic learning tasks: 1. Authentic contexts in which students can experience the way knowledge is used in reality; 2. Authentic tasks and activities in or out of class (e.g., online); 3. Access to real examples of language use (modelling); 4. Multiple roles and perspectives; 5. Collaborative construction of knowledge in and/or out of class (online); 6. Reflection; 7. More explicit knowledge (not only written but also spoken); 8. Teacher's coaching and scaffolding (knowledge building); 9. Authentic assessment (not only tests but also use of presentations, rubrics, flipped class assessment modes and many more).
In connection to the nine elements of authentic learning mentioned above, the study implements more authentic tasks in ESL/EFL teaching and studies their benefits on both knowledge transfer and performance. These effects will be examined in the context of a Task-Based Language Teaching (TBLT) course on ESL/EFL that follows the Content and Language Integrated Learning (CLIL) approach to assure more authentic learning. In the next subsections, the concepts of TBLT, CLIL, and authenticity will be further described. Offering online instruction with more authentic tasks (Tam, 2000) turns learning contexts into more meaningful ones as learners are able to find relevance and value in them.

Task-based language teaching (TBLT)
The TBLT approach to ESL/EFL learning was adopted for this study because, according to Skehan (2003) and Skehan & Foster (2012) and Foster (2009), it places emphasis on both the L2 form and the meaningfulness of learning. Using more holistic learning tasks, which resemble the context of the language application, is considered to more effectively contribute to knowledge transfer. This can be achieved by means of following the three phases of TBLT: 1. the pre-task phase (students are handed out the task worksheet including clear instructions on the activities); 2. the task cycle phase (students decide how to approach the activities with teacher support); and 3. the Language Focus phase (teacher discusses with students language issues, such as the required grammar and vocabulary for task completion).
Compared to more traditional L2 acquisition approaches, such as the Presentation-Practice-Production (PPP) approach, TBLT is said to provide learners with a more holistic view of Second Language (L2) learning. Specifically, the PPP approach to ESL/EFL teaching, as explained by Willis (1996), Skehan (1996a) and Ellis (2003), involves presentation and practice of grammatical items or functions through drill and practice for students to acquire language patterns beyond any real practical context. Within the TBLT approach to ESL/EFL teaching/learning, "Hugging and Bridging" teaching techniques are equally important for knowledge transfer (Authors, submitted a; submitted b). These teaching techniques combine the so-called low road (Hugging) of acquiring new language patterns through drilling and practicing, with the "high road" (Bridging) of acquiring these language patterns through generalizations and active use in real-life situations (Salomon & Perkins, 1989).

Content and language integrated learning (CLIL)
The present study took into account the fact that students' prior knowledge of the English as second/foreign language was used in integration with teaching/learning new content about other domains/subjects. This is directly related to Marsh's (2002) and Coyle's et al. (2010) interpretations of CLIL as an educational approach to ESL/EFL teaching through which students are taught content (school subjects, like geography or biology) in an additional or second language from the one normally used for teaching and learning, English in the case of this study. CLIL places emphasis on both the linguistic demands of a task, like vocabulary and structures, as well as on moving from content to language and back again during the lesson. CLIL lessons for ESL/EFL may relate, as in this study, to teaching the content of Geography, History (Mythology), Environmental Studies, and Mathematics or Science (Physics) subjects. In this way, more experiential learning (TBLT) can be achieved by having trainees complete more holistic tasks, rather than having them practice grammar/syntax items through drilling exercises. To further elaborate on the more experiential aspect of TBLT, Benson (2016) as well as Crookes (1992, 1993) contend that in a task-based approach, students learn language by doing relevant, practical, and engaging tasks which stimulate their interest and sustain their attention, while language is studied on a secondary level as an object necessary for the task completion. Examples of such tasks can be a form of filling out, a reservation making or less complex tasks like food serving.

Task authenticity
Authentic activities are tasks which, according to Ozverir et al. (2017), have real-world relevance and require learners to deeply examine a problem in collaboration with peers, who will propose its solution in the form of a final product to be shared with a broader audience. The study utilizes authentic activities in CLIL lessons which apply both Hugging and Bridging teaching techniques (James, 2006;Perkins & Salomon, 1992;Salomon & Perkins, 1989). Hugging techniques reinforce the acquisition of new language forms by means of creating expectations of how to use new knowledge, connecting it with previous experience, and by using real examples of language. Bridging teaching techniques empower learners by visualizing different contexts in which students learn how to use the new items (generalization mechanisms) in learning to find similar paths for seemingly different problems to solve as well as in learning to keep track (monitoring) of their own ways of thinking (Author et al., submitted a; submitted b). Gulikers et al. (2004) five dimensions of authenticity framework (5DF) has been exploited so that "criteria" or indicators "to determine and describe the degree of authenticity" of the designed tasks (Kim 2007, p. 7) will be ensured. These five criteria used for this study were as follows: 1. the "Task", activity which resembles real life and indicates the way learners integrate knowledge and skills; 2. the "Physical Context", the environment which resembles real life (i.e., an online platform in this study); 3. the "Task Form", the type of assignments to be carried out by students, which includes quality products (e.g., collaborative presentations) that resemble real life (Wiggins, 1989); 4. the "Results/Criteria", being the actual student outcome (i.e., learning performances and the way they were assessed through rubrics in this study); and 5. the "Social Context", the social processes (collaboration/interaction) of students actually working together in groups.

Research questions
In this study, task authenticity was assumed to promote both knowledge transfer and learning performance in the context of an online ESL course. Effects of a "low" and "high" level of authenticity in that course were compared, expecting the course implementing the highest level of authenticity to yield more effective knowledge transfer and learning performance. Low level of authenticity refers to the course (LA) which used L2 form activities like gap-filling, while high level of authenticity (HA) relates to the course which exploited project-based activities. So, the general research question of this study is as follows: (RQ) To what extent does task authenticity affect learning performance and knowledge transfer? From that general question, we can derive three specific research questions: (1) Do high authentic tasks increase knowledge transfer?
(2) Do "high" authentic tasks increase effective learning performance by acquiring teaching strategies? and (3) Are different levels of authenticity provided also perceived and appreciated by students?
The Method section of this study now will describe participants' characteristics, experimental conditions, measurement instruments (ILV, ATI, and 5DF), and the setup of data analysis for finding answers to these questions.

Method
The present study was set up by designing two e-courses that offered L2 activities and tasks (Harmer, 2003) on two distinctly different levels of authenticity, as scored on the dimensions of the 5DF (Gulikers et al., 2006). Questionnaires were answered by 50 students during (ILV) and after (ATI, 5DF) taking the four-week ESL course, and these data could be collected and analyzed to study differences in effects on (perceived) knowledge transfer, learning performance, and perceived authenticity, in order to answer the three research questions.

Participants
Fifty-four students taking English as Second Language (ESL) classes participated in this study that compared two controlled groups of 27 participants. As estimated with the G*Power tool, in order to carry out a one-way ANOVA with two groups and a medium effect size of d = 0,5 (as suggested by Cohen, 1988) and an error margin of 5%, we knew we needed a total sample size of 54 participants for this comparison, with 27 participants in each group. Since the experimental intervention was implemented in two actual classrooms to reach that number, for practical but, most of all, ethical reasons we did not wish to unnecessarily expose even larger numbers of participants, that is, without still knowing if that treatment would work better than existing classroom practice.
Due to a recent national law (2016) about ESL/EFL teaching, Greek students are provided with English instruction in ESL/EFL classes, for which teachers conduct placement tests in the beginning of each new school year. For this placement, three age categories (A: 11-12 years; B: 12-13 years; and C: 14-16 years) and three EQF (European Quality Framework) language mastery levels (A: Beginning; B: Intermediate; and C: Advanced) are distinguished. The four ESL teachers that volunteered to participate in this study chose to work with entering level C students, which are generally 14-15 years of age. Teachers confirmed all participants were indeed in this age range, so we did not have to control for the effects of age and language level in this study. The students were randomly allocated in the two e-courses, yielding two equal and homogeneous groups of 25 participants each. For privacy (General Data Protection Regulation) reasons, data on age and gender of minors could not be directly collected. This is the reason why (trainee) teachers administratively intervened in order to transfer students' answers to the questionnaire items after receiving parents' consent in so doing. Four (trainee) teachers, all women, giving four ESL classes during 4 weeks, were also randomly allocated to the two conditions. Regarding class size, due to the fact that their classes were relatively small, two classes of 14 and 11 students received the same treatment.

Interventions
Both versions of the e-course exploited the CLIL approach to ESL teaching, so taught content (school subjects like Geography, Science, and others) in a different language (English) than the language in which school subjects were usually taught (Greek), but only the experimental version applied TBLT with tasks and teaching techniques providing real and contextualized experiences (Baralt & Gomez, 2017). The following table and description further elaborate the levels of authenticity for each condition, as both e-courses were scored on the 5DF dimensions.
Trainee teachers allocated to the "low" authentic condition, prior to the study, received training on gap-filling, multiple-choice, and matching kind of activity types, as included in the CLIL textbook and used for this version of the e-course. However, they did not use the TBLT approach, which involves more holistic and real-life activities. Trainee teachers allocated to the "high" authentic group received training on more authentic activities exploiting TBLT with tasks as real life-like situations. They, for instance, had to use project-based activities, as included in the CLIL textbook, for that version of the e-course.

Task
The LA course used L2 form activities (e.g., gap filling), while the HA course used project-based activities (e.g., creation of interactive posters by student teams). A more realistic task would be to search online for the political system of a country, depict its basic principles on an interactive poster, and upload it on an e-learning platform.

Physical context
Students in both e-courses performed activities in an online context. Task Form. LA students conducted activities (e.g., matching, multiple choice) for learning new linguistic items (more focus on Hugging techniques), while HA students were taught with more focus on using bridging techniques and produced outputs by means of web 2.0 tools (e.g., shared presentations online).

Results/Criteria
In the LA course, classical forms of assessment were used (e.g., multiple-choice tests), while in the HA course alternative assessment was exploited (e.g., rubrics).

Social context
It was assumed that students would work collaboratively (pairs/groups) in both online courses through the group "breakout mode" of the available e-platform. Instruments (questionnaires). We administered three validated questionnaires (for all scales and their items, see Appendices A, B, and C) to measure the (perceived) effects on knowledge transfer (ILV questionnaire), learning performance (ATI questionnaire), and authenticity (5DF questionnaire).

ILV questionnaire
To measure knowledge transfer, the Integrative Learning Value (ILV) Scale was administered, as was developed and validated by Allen (2014). The ILV contains 12 items (see Appendix A) that are all scored on a 4-point Likert scale, with 0 as the minimum value and 4 as the maximum value. Items are grouped into three subscales: hugging (five items), bridging (four items), and transfer (three items). Since the ILV questionnaire was administered on four different occasions (every week), we collected both weekly scores, average scores, and final delta scores (difference between scores after last week 4 and after first week 1) for the total scale and three subscales. According to James (2006), the items that refer to Hugging are Setting expectations, Matching, Simulating, Modelling, and Problem-based learning) those referring to Bridging are Anticipating applications, Generalizing concepts, Using analogies, Parallel-problem solving, and Meta-cognitive reflection. Those referring to Transfer are Transfer and Connections to other experiences and school domains.

ATI questionnaire
The Approaches to Teaching Inventory (ATI) instrument developed by Trigwell and Prosser (2004) was administered in order to measure learning effects on teaching strategies. The 22 items of the instrument are equally divided over two independent scales: Information Transfer is more teachercentered due to its focus on specific linguistic items with a focus on facts presented to students in order to pass exams; and Conceptual Change can be considered as more student-centered because of its focus on more active involvement and more active ways of thinking through inquiry. The ATI items (see Appendix B) are also measured on a 5-point Likert scale.

5DF questionnaire
The Five-Dimensional Framework of Authenticity (5DF) questionnaire was developed by Gulikers et al. (2006) to measure perceived authenticity of learning activity. It contains 24 items that are measured on a 5-point Likert scale (see Appendix C) and are divided over five independent dimensions of authenticity. As mentioned before, five items indicate the extent to which a task promotes real-life skills, four items the extent to which the physical context in which students conduct a task is made realistic, four items to the extent to which the task form resembles a reallife situation, another group of seven items the results and criteria for evaluation, and a final group of four items the social context in which tasks are collaboratively carried out.

Normality and reliability of scales
Normality of the scales was assessed on the basis of skewness and kurtosis indices, and it was concluded that the data are approximately normally distributed, based on Cameron's (2004, p. 543) argument that "skewness and kurtosis should both fall in the range from+2 to −2 to conclude that data are normally distributed". Table 1 provides statistical descriptives of skewness and kurtosis indices for the dependent variables. Cronbach's Alphas were calculated as a measure of internal consistency of scales. For the ILV scale and its subscales, Cronbach's Alphas were also calculated for each of the 4 weeks, all timepoints yielding excellent values (α > .90). In Table 2, the average alpha values are presented for the ILV scale and its subscales.
It is worth mentioning that the Social Context subscale from the 5DF for Authenticity questionnaire initially produced a negative Cronbach's alpha (α = −1.544), due to a negative average covariance among the four items of the scale. Based on extra tests, item 19 was removed from the scale, and Social Context subscale was again calculated based on three remaining items. This new subscale produced a less questionable Cronbach's alpha (α = .688) which comes close to the .7 threshold. Besides the Physical Context scale of the 5DF, all the other Cronbach's alphas are excellent (with α > .90), and we can safely conclude that all (sub)scales were sufficiently reliable.
Procedure. After the allocation of trainee teachers and their students to conditions, and receiving informed consents from the parents/caretakers of the minor students, teachers delivered their version of the ESL e-course. After each (of four) week, the teachers transferred the answers provided by their students to individual Integrative Learning Value (ILV) questionnaires. After the fifth week, the teachers did the same for the answers their students provided on the ATI and 5DF questionnaires. After data collection, we used SPSS (version 21) software to calculate descriptive statistics (for control of normality and internal consistency of the instruments), to calculate correlations between scales (Pearson's coefficient) and to run analyses of variance (One-Way ANOVA's) for the effects of condition on the outcome variables.

Results
This section presents the analysis of the data collected by the three instruments described in the previous section, in order to provide answers to the three research questions of this study relating the effects of authentic tasks to 1. knowledge transfer, 2. learning performance, and 3. perceived authenticity, respectively.

Knowledge Transfer
In Table 3, the average Knowledge Transfer (ILV) scores are presented per (sub)scale and per week. The total mean scores (over 4 weeks) and the differences (deltas) between (last) week 4 and (first) week 1 were also calculated and presented.
Initially, analyses of co-variance (One-Way ANOVA) were conducted to compare ILV, hugging, bridging, and transfer total mean scores for gender as covariate, with male (n = 26) and female (n = 24) students. We found no statistically significant differences between male and female students neither for ILV (F (1,48) = .344, p = .561), for Hugging (F (1,48) = .424, p = .518), for Bridging (F (1,48) = .258, p = .614), nor for Transfer (F (1,48) = .258, p = .588). Also, gender does not affect Knowledge Transfer growth, as no statistically significant differences were found between male and female students neither for ILV growth (F (1,48) = 1.321, p = .256), hugging growth (F (1,48) = 2.352, p = .132), bridging growth (F (1,48) = .218, p = .642), nor for transfer growth (F (1,48) = .891, p = .350). In Table 4, the Knowledge Transfer (ILV) total mean scores are differentiated by e-course variant (Low or High Authentic). The ANOVA, with experimental condition as factor, indeed demonstrates that Knowledge Transfer is significantly affected by the level of authenticity in the course, as was expected. ILV mean scores are significantly higher in the High Authentic In Table 5, the increase in mean scores on Knowledge Transfer (ILV) and its subscales (hugging, bridging, and transfer) are differentiated by e-course variant (low or high authentic). The ANOVA, with experimental condition as factor, reveals that the effect of condition was not significant, neither for the total knowledge transfer growth nor for two of its three subscales. The only significant difference was found for hugging growth, with F (1,48) = 5.812, p = .020, η 2 =.11, where the LA course demonstrated a higher increase (M = .84, SD = .73) than the HA course (M = .45, SD = .35).
Learning performance on Teaching Strategies. Mean scores on the ATI-scales Information Transfer (M = 31.72, SD = 18.12, n = 50) and Conceptual Change (M = 34.72, SD = 18.53, n = 50) were found to be "moderate" across participants, since the minimum values of these scales are 11 and the maximum values are 55. It is worth noting the existence of a strong negative correlation between both scales (r = −.897, p < .001), indicating that high levels of conceptual change are associated with low levels of information transfer and vice versa. Again, no effect of gender was found on the mean scores on both ATI-scales.
In Table 6 we differentiate the mean scores of these scales for e-course condition (Low or High Authentic). The ANOVA, with experimental condition as factor, shows Information Transfer scores In short, it appears the HA course is more student focused, where the LA course is more teacher focused. This seems to be a quite logical finding since more authentic learning puts students' more active and meaningful learning in focus (conceptual change), where more conventional learning is more teacher-driven (information transfer), see the descriptions of these scales as provided in the previous Method section.
Perception of Authenticity. In Table 7 we present the mean scores encountered on the five dimensions of authenticity over all participants. "Physical Context" shows the highest mean score, while "Task" shows the lowest mean score. Once again, results show that gender does not affect scores on any of the five dimensions.
In Table 8, the mean scores on the five dimensions of authenticity are presented, and differentiated by e-course variant (low or high authentic). ANOVA tests, with experimental condition as factor, were conducted to explore effects of authentic tasks implementation on the five dimensions of authenticity (as perceived by students), which findings follow now at the end of this Results section.  As we expected, students perceived to high authentic (HA) course as more authentic on all dimensions than students in the LA course, however both courses factually only differed on three of these five dimensions (and were similar in physical and social context). "Task" mean scores were found significantly higher for the HA condition

Conclusions
This final section revisits and discusses the main findings obtained from this study that compared two variants of an ESL e-course which varied on their level of authenticity in learning activity. Both courses exploited the CLIL approach to ESL/EFL teaching while making use of Hugging and Bridging techniques within TBLT authentic activities in order to achieve knowledge transfer. The results obtained from this study could confirm most of the expected effects of applying more authentic learning activities in two ESL e-courses, namely Low Authentic/LA course vs. High Authentic/HA course on knowledge transfer, learning performance on teaching strategies and perceived authenticity.
The first hypothesis, that is a positive effect of authentic activities on knowledge transfer, could be confirmed by students' scoring of the ILV scale at four moments (weeks) during the intervention. Specifically, the mean scores of the two ESL e-courses, Low/LA and High/HA authentic, proved to be significantly higher in the HA course indicating that knowledge transfer occurred more effectively when authentic activities (e.g., project-based) were implemented. As stated in the introduction, authentic activities had positive effects on student growth of knowledge and cultural information, also on their effective alignment of L2 learning with real life (Jaccomard & Kuuse, 2016;Ozverir et al., 2017). This becomes evident by their ability to discuss in English more, for example, about the democratic status of various nations after the relevant lesson (HA authentic) having also interacted in their teams on the issue. On the contrary, students of the low (LA) authentic lesson, who focused on the accuracy of the L2 (grammar/vocabulary/syntax activities), could not sufficiently elaborate on the aspects of a modern nation's democratic status.
The second hypothesis, that is, positive effects of authentic activities on both student-focused (conceptual change) and teacher-focused instructional strategies (information transfer) could only be confirmed by scores on the student-focused ATI scale, where results showed that the HA course appeared more student-focused than the LA course. Due to the students' engagement in the lesson (Gilmore, 2007), through their work in groups, presentations on human rights, for instance, were conducted mainly through discussions in which high authentic course/HA students could agree/disagree on the topic. In contrast to that approach, low authentic/LA course students could answer questions about human rights and not actually construct knowledge through personal contributions.
Regarding the third hypothesis, that is, a positive effect of authentic activity on perceived authenticity, all five dimensions of the 5DF framework, namely, the implemented activities, their physical contexts, their forms, their results/criteria, and their social contexts were scored to be significantly higher by students in the High Authentic course. The two examined e-courses, LA & HA, differed in three out of the five dimensions of the exploited (5D) framework of authenticity, namely, the Task (the activity students had to perform), the Task Form (the type of the activity, e.g., multiple choice/poster making), and the results/criteria (what they had to produce, e.g., a collaborative ppt) as expected. That occurred because students in the LA course worked more with L2 form activities, while HA students could create knowledge, for example, through infographics (interactive posters). However, while expected that the rest of the two dimensions, physical context, that is, working in class or online and social context, working individually or in pairs/groups, would be similar in both e-courses, since the lessons were distant through online platforms, the reality was different. The HA course students worked more effectively both in the platform as well as in their pairs/groups than the LA students who did not showcase-efficient collaboration in their teams or adequate presence in the platform as they did not really respond when asked by teachers. This clearly shows that HA course design with more authentic activities motivated students (Buendgens-Kosten, 2014;García-Pinar, 2019) into working more efficiently whether in groups and/or in the online educational platform.
Discussion of research questions. In the context of the above conclusions, the effect found in reply to the first research question of this study further backs up our claim that the implementation of more authentic learning tasks, such as by means of projects, contributes to more effective knowledge transfer (in second language learning) since students appear to work efficiently together in any context. In the same way, students are able to connect the topic under investigation with life-like incidents and make more meaningful contributions whether through discussions or by means of knowledge constructions (interactive creations/syntheses).
Regarding the second research question, the effects found mean that higher learning performance in authentic tasks is encountered in more student-focused (HA) rather than in more teacher-focused teaching/learning (LA) contexts. Specifically, students were driven to more active learning when offered authentic tasks such as activities which required more investigation of the topic (conceptual change), when compared to teacher-focused tasks which focused on transmitting linguistic items to students such as the teaching of vocabulary, grammar, and syntax (Information transfer). After the lessons, students who had worked on tasks were able to converse on various aspects of the studied topic than the students who had only answered questions on the same topic.
Finally, regarding the third research question, the effects found on the five authenticity dimensions (5DF) mean that students perceived and appreciated high authentic tasks more in teaching/ learning contexts that foster holistic, project-based tasks rather than in those employing more grammar/syntax-based activities. Although the physical (class vs. online) and social (individual vs. pair/groupwork) contexts, in which the authentic tasks were provided, were the same for both experimental conditions, it is interesting to observe that even these two dimensions were also perceived and appreciated as more authentic by the respondents (students) who received more authentic tasks.

Limitations.
A number of research limitations should be referred to at this point of the article. In particular, as regards the collection of students' personal data, it was not allowed for them to be openly provided. The ESL teacher trainees could only share them with the researcher after parental consent. In this context, students could not directly fill in the teaching strategies (ATI) questionnaires. As a result of this, the researcher organized an unstructured (free) focus group interview, in the form of an ATI item discussion, to see students' opinions of the teaching strategies implemented during the experiment training e-courses. In this manner, the ESL/EFL teacher trainees made sure that their students had understood the ATI questions so as to monitor them fill in the questionnaire at a later stage. A second limitation was that, instead of one class with 25 students for each experimental condition, four smaller classes of 14 and 11 students participated in the experiment, although all four classes were checked for similarity. Another important limitation seems to be that objective measurements of learning performance and transfer were not objectively obtained from actual task performance, but rather from subjective student perceptions.
Future Research. The results derived from this study can inspire future research in the field of task authenticity, more experiential learning, and meaningful knowledge transfer. This study was aimed at the application of more authentic activities in two CLIL e-courses in relation to the implementation of effective teaching mechanisms for transfer, mainly hugging and bridging (Author et al., submitted b). Besides the authentic tasks per se, opportunities for collaboration and reflection on such tasks for promoting knowledge transfer are also considered as important factors for upcoming research (Radović, Hummel, et al., 2021).
The study context of 14/15 year-olds in ESL-learning activity cannot be automatically generalized towards other domains and school types. Task authenticity can be also located in domains such as pre-primary (nursery) age (Asher & Price, 1967), vocational as well as Tertiary education in which English for Specific Purposes (Kavaliauskienė, 2005) and the CLIL approach to ESL teaching/ learning can be examined. Taking all the above into account, it seems that research into more experiential (more authentic, more reflective, more collaborative) education that fosters meaningful and effective knowledge transfer towards professional practice is a promising endeavor in a broad variety of ESL/EFL teaching/learning contexts to be investigated.

Appendix B. Items of the Approaches to Teaching Inventory (ATI) Scale
Source: Trigwell and Prosser (2004).

2 3 4 5
(1) The student considered important that the (ESL) course would be described in terms of specific objectives.
1 2 3 4 5 (1) In the interactions during the (ESL) course, the student tried to develop a conversation about the topics we were studying. (1) The student had some teaching time to discuss, among themselves, key concepts and ideas when learning English as a Second Language.
1 2 3 4 5 (1) While in the (ESL) course, the student was able to cover the information that might be available from key texts and readings. 1 2 3 4 5 (1) The student was encouraged to restructure his/her existing knowledge in terms of the new way of thinking about a subject invited to develop during the (ESL) course. 1 2 3 4 5 (1) In (ESL) course sessions, the student was provoked to debate and discuss.
1 2 3 4 5 (1) The student was assisted by the teacher in passing the assessment items.
1 2 3 4 5 (1) The student thought it was an important reason for running a(n) (ESL) course if students were handed a good set of notes.
1 2 3 4 5 (1) In (ESL) teaching, the student was provided with the information he/ she needed in order to pass the formal assessments.
(1) During the (ESL) course, the student focused his/her study on what I provided them.

2 3 4 5
(1) The student thinks that the teacher should know the answers to any questions that students may put during (ESL) teaching.
1 2 3 4 5 (1) The student thought that the teacher made available opportunities for students in the (ESL) course to discuss their changing understanding of the subject. (1) The student thinks that it is better for students in (ESL) teaching to generate their own notes rather than copy the teachers.
1 2 3 4 5 (1) The student thinks that a lot of (ESL) teaching time should be used to question students' ideas.
1 2 3 4 5 (1) The student thought that (ESL) teaching focused on the good presentation of information to students.
1 2 3 4 5 (1) The student sees (ESL) teaching as helping him/her develop new ways of thinking in a second language. 1 2 3 4 5 (1) In (ESL) teaching, it is important for the student that the teacher monitors changed understanding of the subject matter. (1) The student thinks that (ESL) teaching focused on delivering what the teacher knew to the students.
1 2 3 4 5 (1) The student thinks that (ESL) Teaching should help students question their own understanding of the subject matter. (1) The student thinks that (ESL) teaching should include helping students find their own learning resources.
1 2 3 4 5 (1) The student thinks that the teacher presented material to enable students to build up an information base in (ESL) learning. 1 2 3 4 5